The problem with TDD as a % (references Khabib vs Conor)

[NotoriousMillions]

Whether or not something is scientific data doesn’t make it automatically true. Scientific data is collected to try and prove(or disprove) a hypothesis. In fact most scientific data has turned out to complete bull crap.

Where is your scientific data? lmao

 

[fortheo]

Debate around the Conor vs Khabib matchup has revealed to me that TDD as % is an unreliable stat when attempting to determine a fighter's propensity to be taken down.

The reason for this is that it doesn't factor in a fighter's ability to control the distance and prevent an opponent from even attempting a takedown. Good strikers often cause good grapplers to be more cautious in their TD attempts to avoid the risk of incurring damage from strikes.

A significantly more useful stat is takedowns allowed per minute. This number uses a fighter's minutes competed and total times taken down to create a stat that shows TD allowed per minute. It captures a fighter's ability to fend off attempted takedowns AND avoid takedown attempts.

Using data from Khabib's UFC opponents, you can see that Conor's TDAPM is better than all but four of Khabib's opponents. If you throw out the Mendes fight (given Conor's knee issues) Conor's ability to avoid the taken is better than all but one of Khabib's opponents.

If my thesis is accurate, you'll see Khabib struggle to secure takedowns on Conor when they fight. I'd love to see data presented on other fighters so please feel free to do some research and share.

Shouldn't you also exclude the khabib fights from all those guys that fought khabib? or even only use their stats from before they fought khabib? that would look at things on a more even playing field since we are comparing them to Conor, who hasn't fought khabib yet.

 

[helltoupee]

Also @helltoupee I don't know if you caught my edit, but I wanted to tell you...good reasonable posts in this thread....

I wish wee could have more genuine honest debate about this fight

Khabib and Conor are both amazing and it's a fascinating matchup

Yet instead sherdog is nothing but Conor haters and "Khabib fanboys" spamming the forums about how Conor sucks and stands no chance

You are one of the few Khabib fans I respect

If I were you, I would hate that people like Mysterio are representing you

I don't know anything about Mysterio, but I do appreciate a civil debate - as you said, both Conor and Khabib are amazing fighters - it should be a very interesting fight - I can legitimately see it going both ways. I can see Khabib ragdolling Conor, and I also wouldn't be surprised if Conor shows some good TDD, strong posts, and lights Khabib up. Conor is very tricky, no doubt, and as many fighters have said, his strength is in hitting you when you don't expect it. That could be a major problem for someone who stands straight up and down, and doesn't use much head movement.

 

[GalegodoVeneno]

I think it was Pettis that met Conor on rehab , he had a bummed knee, why the fuck they would go that far to say that, he won the fight... Nobody need to go flat Earth believers just because of minor shit like that.

 

[VegetaGD]

Am not sure what you want me to consider. I was saying that I didn't see any similarities between your style and Prologic's.

I wasn't addressing you, rather adding to your response to @Mysterio

 

[VegetaGD]

Shouldn't you also exclude the khabib fights from all those guys that fought khabib? or even only use their stats from before they fought khabib? that would look at things on a more even playing field since we are comparing them to Conor, who hasn't fought khabib yet.

I should, 100%

That said, it already took hours to manually collect the data and then coming back and removing the Khabib fight and the minutes those fights took would have been painstaking. I'd love to see someone pickup the torch though because it would be better data.

 

[HIMBOB]

Stats help but context is key.

Take Trujilo. If his grappling wasn't good he wouldn't have been able to be take down so many times because he would have been stuck on his back.

Which makes me think it would be interesting to know how frequently a fighter can return to his feet. Take a guy like Condit stats wise his TDD sucks but he almost always returns to his feet.

 

[fortheo]

I should, 100%

That said, it already took hours to manually collect the data and then coming back and removing the Khabib fight and the minutes those fights took would have been painstaking. I'd love to see someone pickup the torch though because it would be better data.

Also, what matters more than minutes in the UFC in my opinion, would be how many takedowns they actually successfully defended, since that is the real topic up for debate here.

but yeah, i don't expect anyone to put that kind of work in lol

 

[Airdack]

Conor got taken down multiple times by 5'6, 2 week notice Chad Mendes......
He now has to defend takedowns against 5'10, full camp Khabib, who is undefeated and has dominated all of his opponents.....

He got taken down by a 5'6 ncaa all american with the best double at fw. You do know being short and stocky is an advantage for shots? It makes it easier to get under the opponents center.

So yea the 5'6 ncaa all american wrestler who ended up getting popped for roids went 58% on tds on a taller lanky irish striker with supposedly no ground game.

 

[VegetaGD]

Also, what matters more than minutes in the UFC in my opinion, would be how many takedowns they actually successfully defended, since that is the real topic up for debate here.

but yeah, i don't expect anyone to put that kind of work in lol

This is where we diverge. It's about how many times a fighter is taken down. For example, let's say fighter A and fighter B both have 100 minutes of octagon time. Fighter A has defended 8 of 10 takedown attempts (80% TDD) and fighter B has defended 16 of 20 takedown attempts (80% TDD). TDD would tell you that figher A and B are equally capable of avoiding being taken down. Notice that I said avoiding rather than defending. Avoiding encompasses using movement to deter takedown attempts and defending takedown attempts. With the new stat (TDAPM) fighter A is twice as effective as fighter B at avoiding being taken down.

 

[fortheo]

This is where we diverge. It's about how many times a fighter is taken down. For example, let's say fighter A and fighter B both have 100 minutes of octagon time. Fighter A has defended 8 of 10 takedown attempts (80% TDD) and fighter B has defended 16 of 20 takedown attempts (80% TDD). TDD would tell you that figher A and B are equally capable of avoiding being taken down. Notice that I said avoiding rather than defending. Avoiding encompasses using movement to deter takedown attempts and defending takedown attempts. With the new stat (TDAPM) fighter A is twice as effective as fighter B at avoiding being taken down.

Yes, this is where we disagree indeed. You're trying to calculate the strikers ability to use distance control and footwork to avoid shots in the first place, but calculating time in the octagon and comparing it with takedowns defended isn't the way to do it. Placing value on "time in the octagon" in regards to take down defense is flawed because "time in the octagon" does not equal "skills displayed in avoiding or stopping takedowns", therefore when "time in the octagon" increases, it doesn't mean that an ability to prevent takedowns increases or has even been shown at all, assuming it does is just that—an assumption.


For instance, We learn more from fights like Aldo vs. Edgar, where Aldo has to defend takedowns 10-15 times per fight from a guy who we know will actually try to take him down. Conor could have a 200 minute fight against Max where we see zero takedown attempts from Max, a guy who historically doesn't try takedowns, and that 200 minute fight would not tell us more about Conor's Takedown defense than a 25 minute fight with Edgar would have.

That's because time in the octagon isn't necessarily directly related to takedown defense or ability to avoid takedowns in the first place.

 

[Rum]

If you throw out the Mendes fight (given Conor's knee issues) ...

And you were doing so well ...

 

[Alpha_T83]

Conor got taken down multiple times by 5'6, 2 week notice Chad Mendes......
He now has to defend takedowns against 5'10, full camp Khabib, who is undefeated and has dominated all of his opponents.....

1) Conor had the flu and still fought.

2) He had a knee injury and still fought.

3) Why do you fucking morons keep mentioning size. This was 145 Conor against 145 Mendes, so obviously his opponent at 145 is smaller. Conor has put on significant weight since moving up (probably 10-15 pounds). "5'6 Mendes and 5'10 Khabib" is meaningless garbage.

99% of UFC fighters would have pulled out of the fight, after Aldo pulled out. Conor actually showed up and managed to win. People like you that ignore these facts are just blind haters.

 

[Mysterio]

Yes, this is where we disagree indeed. You're trying to calculate the strikers ability to use distance control and footwork to avoid shots in the first place, but calculating time in the octagon and comparing it with takedowns defended isn't the way to do it. Placing value on "time in the octagon" in regards to take down defense is flawed because "time in the octagon" does not equal "skills displayed in avoiding or stopping takedowns", therefore when "time in the octagon" increases, it doesn't mean that an ability to prevent takedowms increases or has even been shown at all, assuming it does is just that—an assumption.


For instance, We learn more from fights like Aldo vs. Edgar, where Aldo has to defend takedowns 10-15 times per fight from a guy who we know will actually try to take him down. Conor could have a 200 minute fight where we see zero takedown attempts from Max, a guy who historically doesn't try takedowns, and that 200 minute fight would not tell us more about Conor's Takedown defense than a 25 minute fight with Edgar would have.

That's because time in the octagon isn't necessarily directly related to takedown defense or ability to avoid takedowns in the first place.

Nailed it.

OP absolutely buried.

 

[bigconan]

Yeah the fight were he was taken down at will by a man half his size until said man gas out because he didn't have a training camp.

first...they weighed the same at weigh ins...= same weight, same size...conor stuffed half on mendes' takedowns on an injured knee...and I have serious doubts about whether he really did have a short camp...anyways, a shortened camp might effect a fighter in the final rounds, but not the 2nd round..he gassed cause of conor's body attack early on.

 

[VegetaGD]

Yes, this is where we disagree indeed. You're trying to calculate the strikers ability to use distance control and footwork to avoid shots in the first place, but calculating time in the octagon and comparing it with takedowns defended isn't the way to do it. Placing value on "time in the octagon" in regards to take down defense is flawed because "time in the octagon" does not equal "skills displayed in avoiding or stopping takedowns", therefore when "time in the octagon" increases, it doesn't mean that an ability to prevent takedowns increases or has even been shown at all, assuming it does is just that—an assumption.


For instance, We learn more from fights like Aldo vs. Edgar, where Aldo has to defend takedowns 10-15 times per fight from a guy who we know will actually try to take him down. Conor could have a 200 minute fight against Max where we see zero takedown attempts from Max, a guy who historically doesn't try takedowns, and that 200 minute fight would not tell us more about Conor's Takedown defense than a 25 minute fight with Edgar would have.

That's because time in the octagon isn't necessarily directly related to takedown defense or ability to avoid takedowns in the first place.

Keep in mind that there is no perfect stat. The point of this thread is that TDD as a % can be very misleading and is inferior to total times taken down. For example, take a look at the TDD defense of the following fighters taken from the UFC website:

• Eddie Alvarez - 93%
• Colby Covington - 78%
• Kevin Lee - 75%
• Conor McGregor - 74%
• Dong Hyun Kim - 72%
• Demian Maia - 66%

The TDD defense stat would lead you to believe that Conor McGregor is significantly better at avoiding takedowns that Demian Maia. But without looking at any data, we'd both know that this doesn't accurately represent who the better defensive grappler is. By utilizing TDAPM, you'd see that Maia over 360 minutes of UFC competition has been taken down 11 times for a TDAPM of .306 almost 2.5 times better than McGregor's .741. Because Maia's grappling offense and control, Demian Maia puts his opponents in a position where a takedown isn't even an option.

I could dive significantly deeper here, but my question (to make sure we're on the same page) is as follows: Are you arguing that TDD as % is superior to TDAPM in assessing a fighter's ability to avoid a takedown? If so, do you agree that Conor is better at avoiding takedowns than DHK and Demian Maia? And Conor about as good as avoiding the takedown as Kevin Lee? Would you pick Eddie Alvarez over Khabib since Khabib would struggle to take Eddie down and wouldn't be able to outstrike him?